Summary Young (1962) has presented evidence which indicates that S associates each item in a serial list with its distinctive temporal position in the list. In this experiment, Young found that items maintaining the same serial position (S-items) in a second list were learned faster than items placed in different positions (D-items), demonstrating a positive effect of position constancy between the two lists. However, it is possible to interpret these results in terms of a chain connecting the S-items with the D-items learned in the original list. The present experiment tested the validity of this explanation in terms of mediation by employing an arrangement of items in which mediation could occur but position associations would not facilitate second-list learning. This was accomplished by shifting the S-items up or down one serial position but maintaining the same relative order in the list. In addition, a further test of the position hypothesis was conducted with a list in which no obvious patterning of S-items was present, allowing the elimination of the mediational chain as a possible aid to second-list learning. The results of the experiment gave no evidence for the operation of the mediational chain either in the learning of the shifted items (Group M) or in the types of errors observed during the second-list learning of Group Y. In addition, the positive effect of constant serial position for the condition (Group P) in which the operation of serial chains was minimized provided further evidence for position association in serial learning.
Read full abstract